⚡ Serverless Image Optimization Pipeline

Automatically compress, resize & convert images to WebP — powered by AWS

Drop a 4MB photo in. Get a 80KB WebP out. Automatically. In under a second.

---

🧩 The Problem

Modern apps are drowning in raw images. Users upload massive smartphone photos — 3 to 8 MB each. Serving those originals to mobile users means:

• 🐢 Slow load times → higher bounce rates
• 💸 Expensive bandwidth bills at scale
• 📱 Terrible experience on 4G/5G networks
• 🔧 Developers manually compressing assets

---

✅ The Solution

A fully serverless, event-driven pipeline that intercepts every upload, optimizes it, and stores a mobile-ready version — with zero human involvement.

Upload JPEG/PNG  →  S3 Trigger  →  Lambda (Pillow)  →  Optimized WebP
     4 MB        →   instant    →   resize + convert  →    ~80 KB

---

🏗️ Architecture

┌─────────────┐     s3:ObjectCreated      ┌──────────────────┐
│   Developer  │ ──── raw-uploads/ ──────▶ │   AWS Lambda     │
│   / App      │                           │   Python 3.13    │
└─────────────┘                           │   + Pillow        │
                                          └────────┬─────────┘
                                                   │
                                          resize + convert
                                          to WebP (quality 80)
                                                   │
                                                   ▼
                                    ┌──────────────────────────┐
                                    │  S3: optimized-delivery/ │
                                    │  filename.webp (~80KB)   │
                                    └──────────────────────────┘

Layer	Service	Purpose
Storage	Amazon S3	Receives uploads, stores output
Compute	AWS Lambda	Runs image processing code
Trigger	S3 Event Notification	Fires Lambda on every upload
IaC	AWS SAM + CloudFormation	Deploys entire stack in one command
Permissions	AWS IAM	Least-privilege read/write access
Image Engine	Pillow (PIL)	Resize, convert, compress

---

✨ Features

• 🚀 Zero infrastructure — fully serverless, scales automatically
• 🔁 Event-driven — triggers instantly on every S3 upload
• 📐 Smart resizing — max 400px width, aspect ratio preserved
• 🖼️ WebP conversion — 25–34% smaller than JPEG at same quality
• 🔒 Secure by default — private bucket, least-privilege IAM
• ♾️ Infinite loop safe — skips already-optimized files
• 📊 Observable — full CloudWatch logging on every invocation

---

📁 Project Structure

mobile-image-optimizer/
│
├── template.yaml              # SAM/CloudFormation — all AWS resources
│
├── image_processor/
│   ├── app.py                 # Lambda handler — core processing logic
│   └── requirements.txt       # Python dependencies (Pillow)
│
└── Serverless_Image_Optimization_Pipeline.pdf   # Full technical documentation

---

🚀 Quick Start

Prerequisites

• AWS CLI configured (aws configure)
• AWS SAM CLI
• Python 3.13

Deploy in 2 commands

sam build
sam deploy --guided

That's it. SAM provisions the S3 bucket, Lambda function, IAM roles, and event wiring automatically.

Test it

# Upload any image to the raw-uploads/ prefix
aws s3 cp your-photo.jpg s3://<your-bucket>/raw-uploads/your-photo.jpg

# Watch the Lambda logs live
aws logs tail /aws/lambda/sam-app-image-optimizer --follow

Check your S3 bucket for the optimized-delivery/ folder — your .webp file will be there within seconds.

---

📊 Real Test Results

Metric	Before	After
File Size	~4,000 KB	~80 KB
Format	JPEG / PNG	WebP
Width	Full resolution	Max 400px
Processing Time	—	< 1 second
Human Effort	Manual	Zero

---

🔮 Future Enhancements

• CloudFront CDN for global edge delivery
• Generate multiple sizes (thumbnail / mobile / desktop)
• SQS Dead-Letter Queue for failed retries
• SNS notifications on completion
• EXIF metadata extraction to DynamoDB

---

💰 Cost

Nearly free for most workloads.

• Lambda — first 1M requests/month are free
• S3 — ~$0.023/GB/month storage
• Example: 100,000 images/month ≈ under $0.25

---

📄 Documentation

A full technical PDF is included in this repo covering architecture, IAM design, data flow, test results, and real-world use cases.

👉 Serverless_Image_Optimization_Pipeline.pdf

---

Built with ❤️ using AWS SAM · Lambda · S3 · Python · Pillow